Fault Tolerant Scheduling of Partitioned and Grouped Jobs in Grid Computing (FTPG)

Abstract

Computational grids have the potential for solving scientific and large - scale problems using heterogeneous and geographically distributed resources In addition to the challenges of managing and scheduling resources reliable challenges arise because the grid infrastructure is unreliable There are two major problems in Scheduling the Grid 1 Efficient Scheduling of jobs 2 Providing fault tolerance in a reliable manner Most of the existing strategies do not provide fault tolerance There are some algorithms which provide fault tolerance but they do a large amount of redundant computation to provide fault tolerance This paper addresses this issue and minimizes redundant work by using a group level table of data This technique is suitable for partitioning and group scheduling of jobs